69 results about "Semantic computing" patented technology

The invention provides a semantic computing method for improving a word vector model. The method comprises the following steps of S1, preprocessing a corpus; S2, tagging parts of speech and tagging the parts of speech of words obtained through preprocessing of the corpus; S3, initializing vectors and vectoring the words obtained through tagging of the parts of speech and the parts of speech; S4, integrating context vectors and computing and integrating context word vectors and part of speed vectors of the words; S5, establishing a Huffman tree, training a network, optimizing a target function and judging whether an error reaches a threshold value or not; S6, obtaining the vectors and obtaining the word vectors and the part of speed vectors; and S7, applying the vectors and carrying out semantic computing through application of the word vectors and the part of speed vectors. Compared with the prior art, the method has the advantages that a part of speech factor is added to the vectors, an existing Word2vec model is improved, moreover, innovative application is carried out according to the improved model, and a function of carrying out semantic computing through the Word2vec is expanded.

The invention discloses a mass disturbance warning method and system applied to short texts. The method comprises the following steps of automatically structuring a domain knowledge database applied to mass disturbances; combined with the domain knowledge database, performing local structured extraction and online classification on the short texts to screen out mass disturbance texts with potential risks from mass short texts; combined with the domain knowledge database, performing overall structured extraction and online clustering on identified short texts, and according to whether the number of short texts contained in every cluster exceeds a preset threshold value, determining whether to perform timely alarming. The mass disturbance warning method applied to the short texts has the advantages of fully integrating domain background knowledge, context, shallow semantic expression and deep semantic computation, achieving collaborative analysis and predication of mass disturbances and improving the timeliness, accuracy and recall rate of screening, tracking and alarming of the mass disturbances.

Task Computing computer system by segmenting the system into a plurality of implementation tiers of a presentation layer, a remote procedure call programming interface (API), a middleware layer to which the presentation layer interfaces via the remote procedure call API to real-time, dynamically generate a computer implemented task interface at the presentation layer to a semantically described source of function as a service on a computer system, and a service layer and a function source realization layer providing the semantically described source of function as the service on the computer system to which the middleware layer interfaces. Real-time and dynamically composing an executable task that comprises one or more services using the generated task interface at the presentation layer to one or more services on the computer based upon the semantically described application-, device- and service-rich computer.

The invention discloses a man-machine interactive system and method for digital television voice recognition. The system comprises a target voice acquisition module, a voice analysis module, a semantic computation module and an intelligent control module, wherein the target voice acquisition module comprises a signal amplification module, a forward filtering module, a signal sampling module and a data compression and coding module; and the voice analysis module comprises a noise removal module, a feature extraction module and a decoding device. The method comprises the processes of target voice acquisition, voice noise removal, voice recognition processing, command recognition conversion and intelligent control processing. In the invention, through the cooperative work of the modules, a digital TV man-machine interactive technology for anti-interference voice intelligent identification and voice analysis and interaction under the digital TV reverberation acoustic environment of digital home life is achieved, and an advanced digital TV voice language interaction mode is provided.

The invention discloses a method and a system for excavating information of academic journal paper authors. The method comprises the steps: firstly selecting an objective subject field and establishing an OWL (Web Ontology Language) field ontology; secondly extracting author information from an academic journal paper of the objective subject field; thirdly performing the format conversion to the extracted author information, storing the information into an author information bank and calculating out the unique author ID (Identity); finally obtaining an incidence matrix of an author and an academic paper, an academic developing route chart of the author, a cooperator network diagram of the author, an academic cooperation distances among authors, a hot spot research direction map and an academic reputation map of the author by above information. According to the invention, the data source of the author information excavating method is changed; the OWL filed ontology technology is introduced into calculation processes of the academic cooperation distances among authors and the hot spot research direction; and the semantic calculation effect is improved.

The invention relates to an automatic abstract generation method based on a concept semantic unit. The method comprises the steps that the concept semantic unit is utilized to serve as a carrier of semantic computation and document content representation; a semantic focus of document content is obtained through convergence of document semantic content; automatic generation of a document abstract formed by selecting abstract sentences representing the document content according to the semantic focus is achieved. According to the automatic abstract generation method based on the concept semantic unit, by means of a potential Latent Dirichlet Allocation (LDA) mode, modeling is conducted on a document theme, theme generation treatment is achieved, and the abstract sentences are selected according to the importance degree of the theme. In the computation process, a concept hierarchy network symbol is introduced, merging to semantic information is achieved by means of semantic hierarchy relations in the concept hierarchy network symbol, and the data sparsity problem caused by taking a word as a semantic computation unit is improved.

The invention discloses a specialized vocabulary meter dynamic generating system, which is characterized by the following: basing on theme semantic vector; proceeding professional sort or cluster document sort module and specialized vocabulary meter generating module for correspond document; extracting finite quantity and a professional sort document text with the specialized vocabulary meter; calculating weight for all vocabulary in the text; ordering with the size of the weight value; choosing the front N vocabulary as specialized vocabulary meter of relative professional sort. This invention also disclose a specialized vocabulary meter dynamic generating method. This invention possesses high accuracy and low cost, which can proceed dynamic updating maintenance for specialized vocabulary meter.

The invention relates to a semantic-computing-based interaction system and a semantic-computing-based interaction method for person with hearing or language disorder. The system is characterized by comprising two parts, namely portable terminals and a cloud-computing-based cloud server, and comprises at least two types of portable terminals, namely a first-type portable terminal and a second-type portable terminal or a portable terminal combining the two types of terminals. The portable terminals are used for performing related computing, data also can be uploaded to the cloud server for accurate computing, then a result is returned to the portable terminals, and different users can customize different semantic dictionaries. The system and the method have the advantages that words spoken by normal person are converted into the animation display of hand signs, so that the normal person can transmit information to the person with the hearing or language disorder without barriers; and the hand signs of the person with the hearing or language disorder can be converted into voices for play or into texts for display, so that interaction between the normal person and the person with the hearing or language disorder is further facilitated.

The invention provides a computing method, a search processing method, a computing device and a search processing device for sentence similarity realized through a computer. The computing method comprises the following steps: obtaining a first sentence and a second sentence; carrying out dependency analysis on the first sentence and the second sentence respectively to obtain a first dependency tree and a second dependency; computing the sentence similarity of the first sentence with the second sentence according to the first dependency tree and the second dependency tree. The search processing method comprises the following steps: receiving a query sentence; obtaining at least one search result item according to the query sentence; respectively computing the semantic similarity of the query sentence and the semantic similarity of the search result items through the computing method for sentence similarity realized through the computer; ordering the search result items according to the computed value of the semantic similarity; sending the search result items through ordering. More accurate sentence similarity can be computed according to the semanteme of sentences, and the search results are more accurate.

The invention relates to a method for information recommendation in a could environment based on data semantics. The method achieves the information recommendation through a basic data and user preference information semantization module, a could environment mass semantic information indexing module and a semantic calculation based information recommendation module to achieve the information recommendation, wherein the basic data and user preference information semantization module obtains basic data and user preference information through a cloud platform and conducts semantization description on the basic data and the user preference information to form a basic data and user preference information body library, the could environment mass semantic information indexing module establishes an indexing structure for semantic information and performs index division and regrouping when indexing nodes are overload, the semantic calculation based information recommendation module conducts semantic calculation on the basic data and user preference information body to obtain an information recommendation result. Compared with the prior art, the method for information recommendation in the could environment based on data semantics has the advantages of being good in real-timeliness, high in robustness and recommendation quality and the like.

The invention belongs to the technical field of information retrieval and database structure, discloses a classifier construction method based on semantic computation, and a classifier. A neural network model is used to train word vectors on Wikipedia data to obtain distributed representation of words; vector representation of a classification is obtained through a classification label, and a weighted average method is used to obtain vector representation of a text; and by calculating the semantic relationship between a classification vector and a text vector, the most likely classification towhich the text belongs is obtained. The unsupervised learning phase does not need to data labeling, and only needs a user to define a small number of feature words to complete the creation of the classifier, the online speed is fast, and there is no need to wait for long label data accumulation; the unsupervised learning stage can make full use of the existing limited label data, and guide and improve unsupervised classification by extracting valid feature words.

The invention provides an intelligent retrieval method and device for calculating patent literature similarity based on word frequency and semanteme, electronic equipment and a storage medium of the electronic equipment. Bag-of-words statistics and word vector calculation are conducted on all literatures in a patent database, and corresponding bag-of-words data and word distance data are obtained;the method comprises the following steps: establishing a model, inputting contents or examination question numbers, acquiring titles, abstracts, claims and specifications of patents to be examined from question bank data and carrying out various combinations, performing rough selection and fine selection respectively according to a bag-of-words algorithm and a semantic algorithm, performing textsimilarity analysis on selected data, and performing fusion sorting on analysis results to obtain comprehensive similarity. Through duplicate checking and screening, a suspicious answer set of the to-be-checked patent is given. According to the method, the retrieval speed is increased, two rounds of screening are adopted, the first round of roughing aims at rapidly narrowing the comparison range,and the second round of fine selection aims at improving the accuracy; manpower and time can be effectively saved, a patent reviewer is helped to reduce the related patent review range, and review efficiency is improved.

The invention discloses an intelligent workshop multi-source heterogeneous manufacturing big data integration model and a semantic calculation method. According to the method, a first procedure servesas a root node, and a processing time sequence flow serves as a model stem of the model in an evolution direction; all processing procedures serve as branch nodes, resources relevant to the procedures are associated, branches are formed through derivation, and real-time data obtained in the manufacturing process serves as leaf nodes of the intelligent workshop multi-source heterogeneous manufacturing big data integration model to establish the model; and comprehensive processing progress information, equipment load information and product circulation and logistics trolley transportation trackinformation of products in a workshop are obtained through a calculation formula, and description, association, fusion, mining, semantic calculation and knowledge evolution of multi-source heterogeneous manufacturing big data are realized through the dimensions, granularity, mass and other aspects of the associated manufacturing data.

The invention discloses a personalized multi-view federal recommendation system, which comprises a central server and a plurality of user clients, and any user client comprises a training module and a prediction module; wherein the training module comprises a data distribution sub-module, a gradient calculation sub-module, a gradient aggregation sub-module, a model updating sub-module, a model fine tuning sub-module, a user data warehouse and an article data warehouse which cooperate with one another to complete execution of a training algorithm, and a user sub-model and an article sub-model are obtained; and the prediction module comprises a semantic calculation sub-module, an interactive calculation sub-module, a probability aggregation sub-module, a probability sorting sub-module, a recommendation output sub-module, a user model warehouse and an article model warehouse which cooperate with one another to complete execution of a prediction algorithm and obtain a recommended article sequence corresponding to any user client. According to the method, the scene adaptability is higher, the feature mining of the underlying model is deeper, the data source covered by the original input is wider, and the localization fine tuning of the global model is better.

The invention provides a bandwidth-saving video transmission method. According to the method, video data are deconstructed into text semantic information at a sending terminal to realize textualized transmission of key information; and the text semantic information is constructed into a reconstructed image consistent with the semantic information of an original target image at a receiving terminalbased on an image generation technology. Therefore, different forms of computing capabilities are given to each part in the communication network structure; specifically, a mode of endowing a sendingterminal with a video data feature semantic computing capability and endowing a receiving terminal with a semantic generation image computing capability is adopted, and an original local centralizedcomputing communication mode is converted into a cooperative diversity computing communication mode, so that the bandwidth and storage resources are greatly reduced, and the transmission, storage andretrieval efficiency is improved.

The invention provides a method for generating animation head leader captions and tail leader captions, which comprises the following steps: 00) establishing a NarLog caption language source file of which each line comprises an attribute name and an attribute value; 10) performing syntactic analysis on each line of the Narlog caption language source file to identify the attribute type of the line; 20) if the attribute type is a display attribute, performing semantic computation on an animation attribute under the current context according to the display attribute of the current context so as to generate a script subroutine, deleting the animation attribute under the current context, and using the display attribute to update the context; 30) if the attribute type is the animation attribute, adding the animation attribute to the current context; and 40) making use of a Maya script editor to execute the generated subroutine so as to generate the captions through rendering. The method can solve the problem that handmade animations consume more manpower and financial resources, and can generate the animation head leader captions or the tail leader captions automatically.

The invention provides a semantic recognition method based on multistage semantic representation and semantic calculation, and mainly solves the problems of poor interpretability and low generalization ability of signal semantic recognition in the prior art. The implementation scheme of the method comprises the steps of collecting a training set and a test set; constructing a signal semantic recognition network composed of cascaded multistage semantic representation networks and semantic computing networks so as to carry out learnable multistage semantic representation on the signals, and computing semantic categories of the signals according to the semantic representation; setting a semantic representation loss function and a cross entropy loss function to train the multistage semantic representation network and the semantic computing network in sequence to obtain a trained signal semantic recognition network; and obtaining a semantic recognition result of the to-be-recognized signalbased on the trained signal semantic recognition network. According to the method, the interpretability and generalization ability of signal semantic recognition are effectively improved. The method can be used for man-machine interaction and semantic information retrieval.

The invention relates to an artificial intelligence technology, and discloses a report generation method, which comprises the following steps: acquiring a configuration demand of a user, and extracting demand semantics of the configuration demand; calculating a matching value between the demand semantics and a configuration label of each preset configuration in a plurality of preset configurations, and selecting the preset configuration corresponding to the configuration label with the maximum matching value as a target configuration; obtaining a data list initial framework, and configuring the data list initial framework according to the target configuration to obtain a target framework; and performing database connection configuration on the target framework to obtain a data list. In addition, the invention also relates to a block chain technology, and the data list can be stored in the nodes of the block chain. The invention further provides a report generation device, electronic equipment and a storage medium. The report generation efficiency can be improved.

The invention particularly relates to a travel intention prediction method based on a double-attention graph embedded network. The travel intention prediction method comprises the following steps: establishing a travel intention prediction model; obtaining travel track data of the user and corresponding interest point sign-in data; the method comprises the following steps: firstly, aggregating travel trajectory data and interest point sign-in data through a travel intention prediction model to enhance activity semantics corresponding to a time-space situation, a starting point interest point situation and an end point interest point situation; embedding comprehensive activity semantics of the travel intention extracted in the advanced feature space based on the space-time situation, the starting point interest point situation and the terminal point interest point situation through the double attention graph; finally, the prediction probability of each candidate activity is calculated based on the comprehensive activity semantics of the travel intention; and taking the candidate activity with the highest prediction probability as the predicted travel intention of the user. According to the travel intention prediction method, personal sensitive information of the user does not need to be obtained, so that the privacy and practicability of travel intention prediction can be improved.

The invention discloses a mapping method from WordNet to Neo4J, a semantic detection method and a semantic calculation expansion interface generation method, and the method comprises the steps: extracting effective information of a synonym set, and mapping the effective information to synonym set nodes in the Neo4J; adopting a breadth-first algorithm to traverse each two synonym collection points associated with each other according to a specific relationship, and generating a connection between the corresponding Neo4J synonym collection points; traversing all words in the WordNet and the semantic element set corresponding to each word, and binding the words to the synonym set based on the semantic element set to form word nodes. According to the method, high-efficiency semantic calculation is realized by adopting the WordNet-based semantic relation network, the realization of semantic calculation floats to a Cyper language layer, the expression is more visual, the semantic calculation is visualized, the semantic calculation is mapped to a graph structure of Neo4j, and the data processing speed is higher. The optimization of the graph structure of the Neo4j enables the query in the implementation process of the semantic detection algorithm to be reduced from the spatial complexity On^n to O1.

The invention discloses a method and a device for the positive and negative analysis of Chinese comments, aiming at solving the problem of the inferior accuracy of analytical results when Chinese comments are analyzed by adopting a semantic computation-based method in the prior art. The method comprises the steps: translating the input Chinese comments into corresponding English comments; splitting the English comments resulted from translation, matching the splitting result with corresponding English semantic resources in English positive/negative glossary and implementing positive and negative analysis; and outputting the analytical results as a positive/negative tendency value for the implementation of positive and negative analysis on the Chinese comments.

The invention discloses a BERT-TextCNN-based complex equipment manufacturing BOM intelligent decomposition method, which comprises the following steps of: firstly, performing text preprocessing on a manufacturing BOM through three steps of category labeling, semantic expansion and stop word removal; secondly, performing vectorization description on the preprocessed BOM text statement through an embedding marking method, an MLM mechanism and an NSP mechanism; and finally, performing feature extraction and classification on the vectorized BOM text statements by utilizing a BERTTextCNN, and realizing intelligent decomposition of the complex equipment manufacturing BOM according to four types of necessary self-made parts, necessary outsourcing parts, non-necessary self-made parts and non-necessary outsourcing parts. The method has the characteristics of low cost and high efficiency.

The invention discloses a space querying method based on an interaction-type sketch. The method is suitable for use in electronic map-oriented fuzzy scene querying, solves contradiction between increasingly-rich geographical data and a single current space querying mode and result display form, and includes the following steps: (1) providing a space querying-oriented sketch scene construction environment, in which elements can be drawn through strokes in a drawing process, and forms of tags and the like can be utilized to input attribute information of the elements; (2) calculating space relationships between the sketch elements; (3) constructing a semantic graph of a sketch scene; (4) calculating scene similarity on the basis of semantics; and (5) outputting similar space scene querying results. According to the method, the interaction-type sketch is applied to space querying, multi-element scene matching is converted into semantic similarity degree calculation, and modes of geographical information retrieval can be effectively expanded.